In an attempt to identify new targets for immunotherapy and to investigate whether endogenous immune responses might already exist in patients with Ewing's sarcoma, we have conducted a series of experiments to study whether endogenous T cell mediated immunity directed toward autologous tumors are present in patients presenting with Ewing's sarcoma. Autologous T cells collected by apheresis from patients with Ewing's sarcoma were tested for reactivity toward autologous tumors. These studies showed that cytolytic T cells circulate in patients with Ewing's sarcoma. These tumor reactive killers can be expanded and activated by CD40L-matured DCs following by re-stimulation with tumor cells, or anti-CD3 and 4-1BB antibodies in the absence of autologous tumor. Using these methods, we generated tumor-lytic T cell responses in 6/6 patients. Interestingly, tumor lysis not consistently MHC restricted, whereas tumor induced proliferation showed consistent MHC restriction. We identified activation NK receptors on the lytic T cells, which provided a basis for the promiscuous cytotoxicity. Phenotypically, the tumor reactive cells were CD3+CD8+CD28- and the subset which produced IFNgamma in response to autologous tumor were CD8+/CD28-/4-1BB+ (CD137)+ whereas CD8+/CD28-/4-1BB- (CD137-) cells do not. Therefore, 4-1BB expression is an important marker for the tumor reactive subset in Ewings sarcoma. 4-1BB is a member of the tumor necrosis factor receptor (TNFR) superfamily that is preferentially expressed on the surface of CD8+ T cells after antigen-induced activation and cross-linking of 4-1BB and T cell receptor (TCR) on activated T cells has been shown to deliver a costimulatory signal to T cells. Thus, during the course of these studies we investigated whether Ewing's sarcoma itself might be capable of providing the 4-1BB mediated costimulatory signal. We discovered that Ewing's sarcoma cell lines themselves express 4-1BB ligand. Furthermore, T cell activation could be induced by the tumor alone but required both signal one (presumably provided by a tumor associated antigen) and signal two which is provided by 4-1BB ligand expressed by the tumor. When either of these interactions are blocked, the tumor induced proliferation of autologous T cells is blocked. Finally, using a SCID-hu PBL model wherein autologous Ewings tumors are xenografted onto the extremity of mice, we observed diminished growth of primary tumors and diminished metastatic disease when anti-CD3/4-1BBL tumors we administered compared to anti-CD3/anti-CD28 or compared to administration of non-stimulated CD8+ T cells. These results provide the first evidence that endogenous immune responses exist in patients with Ewing's sarcoma and question the widely held view that progressive tumor growth induces T cell tolerance. Furthermore, these results open the intriguing possibility that tumor expression of an immune mediated costimulatory molecule may provide ongoing signals which sustain an immune response in vivo. We are currently optimizing approaches to expand the CD3+CD8+4-1BB+ subset from both normal hosts and tumor bearing hosts and can currently accomplish a two log expansion of these cells. These results have been recently published (Zhang et al., 2003). We are also currently studying the biology of programmed cell death in Ewings sarcoma in an effort to identify potential blocks in death pathways which may modulate susceptibility to immune based therapies. Previously, our group published evidence for susceptibility to Fas mediated cell in Ewings sarcoma and more recently published that Ewing's sarcoma is exquisitely susceptible to TRAIL mediated cell death in vitro. Current studies are underway to determine whether this susceptibility in vitro translates into efficacy of TRAIL based therapies in vivo using murine xenografts. These studies have identified activity of TRAIL receptor ligating therapies in vivo, but the potency of the response is modest compared to the exquisite sensitively observed in vitro. We have thus undertaken studies to characterize the relative TRAIL resistance which is induced in vivo. Interestingly, we observe TRAIL resistance of tumor explants during the first several passages ex vivo of tumor derived from either TRAIL agonist treated on non-treated animals. This resistance is characterized by reversal using IFN gamma as well as downregulation of TR2. When TRAIL receptor agonists are combined with IFN gamma in vivo, we observe an improved capacity for TRAIL based therapy to prevent metastatic disease. Through a CRADA with Human Genome Sciences, we are currently planning to develop a Phase I trial of anti-TR1 and anti-TR2 monoclonal antibodies in pediatric solid tumors.